No label defined (Q53685)

An old infill idea of mine that uses rhombic dodecahedra. It's intended to be a low density fill using walls that are 2 paths wide. Walls are vertical or at 45 degrees. The walls being angled help evenly spred forces, so this should be quite strong, although with a large cell size you may want to additionally fill the outermost cells with a finer resolution infill. The paths look like they have different spacing, but they are the same when you account for slope. The filament paths cross over each other, providing extra strength binding the layers together better. This way it knits the layers together as filament is pushed solidly onto the layer below and filament beside. If there is extra plastic, this effect is even stronger. I am also considering simultaneously moving the z axis to enforce this. This crossing over also keeps the filament from being pulled into the gaps when cornering. You could also vary how much material is being put down if there is too much plastic buildup, i.e. reducing flow when it crosses. There are not sharp turns (all greater then 90 degrees) so head velocity can be kept up. Pictures of the test objects are just the shape of the rhombic dodecahedral infill put threw skeinforge. They don't employ this algorithm. Thats why they look so messy. This is from January 2011, although I've been thinking about it for years... at least 2009. I should probably post stuff more often. Anyway please comment and discuss about this. I love positive and negative feedback.